I have an emotion database of 213 (with 7 classes). I used a bank of Gabor filters for the extraction of features. So I got a data matrix of 213x50000 (a huge number of parameters on only 213 images !!). So i have to reduce this large number of features before learning with SVM, and i decided to reduce with PCA. Is it a good choice?

  • $\begingroup$ use Convolutional neural network instead of PCA + SVM $\endgroup$ – Haitao Du Apr 20 '20 at 12:45
  • $\begingroup$ but i would used Gabor filter not deep learning approach $\endgroup$ – Adel Madrid Apr 20 '20 at 13:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.